An integrated platform for bovine DNA methylome analysis suitable for small samples
© Shojaei Saadi et al.; licensee BioMed Central Ltd. 2014
Received: 17 February 2014
Accepted: 28 May 2014
Published: 9 June 2014
Oocytes and early embryos contain minute amounts of DNA, RNA and proteins, making the study of early mammalian development highly challenging. The study of the embryo epigenome, in particular the DNA methylome, has been made accessible thanks to the possibility of amplifying specific sequences according to their initial methylation status. This paper describes a novel platform dedicated to the genome-wide study of bovine DNA methylation, including a complete pipeline for data analysis and visualization. The platform allows processing and integrating of DNA methylome and transcriptome data from the same sample. Procedures were optimized for genome-wide analysis of 10 ng of DNA (10 bovine blastocysts). Bovine sperm and blastocysts were compared as a test of platform capability.
The hypermethylation of bovine sperm DNA compared to the embryo genome was confirmed. Differentially methylated regions were distributed across various classes of bovine sperm genomic feature including primarily promoter, intronic and exonic regions, non-CpG-island regions (shore, shelf and open-sea) and CpG islands with low-to-intermediate CpG density. The blastocyst genome bore more methylation marks than sperm DNA only in CpG islands with high CpG density. Long-terminal-repeat retrotransposons (LTR), LINE and SINE were more methylated in sperm DNA, as were low-complexity repetitive elements in blastocysts.
This is the first early embryo compatible genome-wide epigenetics platform for bovine. Such platforms should improve the study of the potential epigenetic risks of assisted reproductive technologies (ART), the establishment sequence of embryonic cell lines and potential deviations in both gene expression and DNA methylation capable of having long-term impact.
KeywordsEpigenome DNA methylation Bovine embryo Methylome and transcriptome parallel analysis Analysis pipeline CpG enrichment Repetitive elements Epigenome-wide association study
The study of early embryonic development continues to pose formidable technical challenges due in large part to the limited amounts of sample material. However, high-throughput high-fidelity amplification of nucleic acid is making the macromolecular study of embryonic physiology more accessible. Microarray platforms, and more recently RNAseq, have made studying the early embryo transcriptome almost routine. Our group has been developing bovine and porcine microarrays based on transcriptomic platforms that include standardized sample preparation procedures and a complete user-friendly software suite for data normalization and analysis, allowing efficient processing of samples from extraction through to the generation of publishable graphs [1, 2]. Transcriptomic platforms have been used to study how early embryos of many different species interact with their immediate microenvironment [3–8]. Although very useful, the transcriptome has not allowed us to determine whether or not deviant gene expression, that is observed in embryos, is a transient adaptation to surrounding conditions that later yields to normal expression without any long-term impact on development. In order to provide a more complete picture of embryo adaptation and its potential long-term consequences, study of the epigenome is necessary [9, 10]. The epigenome is the sum of all epigenetic information  and refers more precisely to the complete description of chemical changes to DNA and histones , including histone tail modifications, chromatin remodelling proteins, and ncRNA. Epigenomics and transcriptomics are closely interrelated in terms of gene function and regulation [13, 14] and together modulate gene expression.
Among epigenomic effectors, DNA methylation is believed to be a strong primary molecular mark having a major impact on intergenerational gene silencing [15, 16]. DNA methylation patterns are known to be relatively stable and established in a tissue-specific manner [17, 18]. However, following fertilization and during mammalian pre-implantation development, the DNA methylation pattern is dynamic and undergoes reprogramming in the form of a wave of genome-wide de-methylation and re-methylation [19–24], thus putting the embryo at risk of programming errors . Furthermore, the study of how the DNA methylome can be modified by changes in the embryo microenvironment such as in vivo/vitro culture, uterine conditions, or maternal nutritional regimen has represented a major challenge and continues to do so, due mainly to sample scarcity offering input DNA well below minimal recommendations.
Numerous platforms already exist to study methylation of targeted loci or to obtain genome-wide methylation profiles. For the study of very small samples, determining DNA methylation at targeted loci has so far been more successful than genome-wide approaches . The main advantages of general survey are the possibility of describing physiological responses at the genome-wide scale and the potential for novel discovery.
The aim of the present work was to develop a technological platform that is complementary to existing platforms, in order to provide a whole-genome view of DNA methylation in bovine early embryos. Since a diploid mammalian nucleus contains about 6.8 picograms of DNA (http://www.genomesize.com) and the expanded blastocyst of large mammalian species is composed of about 150 cells, a single bovine blastocyst contains approximately 1 ng of DNA. The current benchmark for minimal sample size is around 10 ng, therefore corresponding to a pool of 10 expanded blastocysts. The other criterion is ease of use, at both the sample handling and data processing steps. We thus sought to identify an existing methodological approach which would be best suited to analyze very small samples of DNA. The platform was tested and validated using experimental samples.
The analysis pipeline
DNA methylation profiles of bovine sperm and early embryos
Genome-wide overview of DNA methylation profiles.
Characteristics of differentially methylated regions (DMRs).
Development of EDMA
The study of the epigenome has become a prime focus in the effort to understand the expression of complex phenotypes such as obesity, diabetes [29, 30], mental disorders [31, 32] and cancers [33–36]. Vast epigenetic erasure and reprogramming events during early embryogenesis make this phase of development a window for perturbations, potentially having long-term impact on phenotype in adulthood. Methods for monitoring candidate loci using as little as single cell , one oocyte  or single cell blastomere  have been developed. The aim of the present work is to develop a robust technological platform to study genome-wide DNA methylation, suitable for the small amount of sample material obtainable from early embryos and to complement this platform with a comprehensive suite of tools for quality control and data analysis. The EmbryoGENE DNA methylation analysis (EDMA) platform was developed for compatibility with our previously published transcriptome analysis platform  with the aim of integrating data of both types from the same sample.
Among the three sample treatment options currently employed to study genomic DNA methylation , we opted for methyl-sensitive restriction enzymes (MSREs) over bisulfite conversion and affinity selection because it allowed for robust processing of DNA input of a few nanograms as well as a more straightforward data processing based on the well-established tools dedicated to microarray data analysis. An optimal procedure was sought for isolating intact genomic DNA for subsequent specific fragmentation by nucleases, which offers better repeatability than mechanical shearing (data not shown). The selection of methylated fragments thus obtained is performed by means of successive PCR reactions, while amplicon identification is achieved using a microarray, which brings the added benefit of mature data processing and analysis procedures. In addition, each microarray slide is composed of two arrays that hold two samples each (two-color arrays), allowing for the necessary inclusion of biological replicates in the experimental design. All platforms must include biological replication, since the extent of DNA methylation is naturally variable among samples of similar origin.
The procedures involved in EDMA, namely restriction endonuclease genomic fragmentation, adaptor ligation, cleavage by MSREs, selection of protected fragments by PCR and identification of fragments by microarray, are essentially those proposed for nanoHELP and for array-comparative genomic hybridization, which have been applied successfully to inputs of respectively 10 ng and a single cell . Ten nanograms is the approximate quantity of DNA obtainable from 10 expanded blastocysts. Comparing 1 μg and 10 ng genomic DNA input, Oda and colleagues demonstrated that the smaller input provided higher reproducibility (R > 0.96), due probably to more complete cleavage, suggesting that low input of genomic DNA would likely be more suitable for MSRE-based platforms . Sample processing used in EDMA was tested with lower gDNA inputs as low as 15 cells and genomic coverage was measured using the 50 K bovine SNP Chip which showed important loss of information due to allele drop-out when input were lower than 1 ng (data not shown). When dealing with single cell DNA methylation analysis, partial sample loss is often observed. This loss may not be problematic when aiming for targeted loci but is definitely problematic for genome-wide profiling. As such, 10 ng was set the minimal input leading to a robust DNA methylation profiling.
In order to prevent the introduction of false positives that would arise from incomplete cleavage of sample DNA [42, 43], spiked-in controls were designed to account for extent of cleavage both during the initial MseI fragmentation and by MSREs. Samples were tested before fragment amplification by PCR to ensure that cleavage was sufficient. When cleavage is incomplete, the MSREs treatment can be repeated. For each MSRE, a pair of synthetic DNA controls was designed, one of which was methylated in vitro and used as a control for fragment protection, while the unprotected counterpart acted as the cleavage target.
One of the main concerns with any genome-wide approach is the actual genomic coverage. Since the MseI recognition site is T/TAA, CpG dinucleotides are left intact and most CpG islands are conserved . These sites are very abundant throughout the genome, resulting in small fragments well suited for LM-PCR [40, 45, 46]. By fragmenting the entire genome in this manner, methylation status outside CpG islands can be queried. This provides valuable information, since evidence is mounting for potential important roles for 5mC in non-CpG islands, which have been found more dynamic than CpG islands with respect to methylation-based regulation . This is potentially even more important in early embryos where overall demethylation just occurred.
Since detection based on restriction enzymes is limited to fragments bearing the recognition sites [11, 48], adequate genomic coverage by MSRE requires a combination of nucleases [27, 44]. In silico analysis showed that compared to HpaII alone, the combination of HpaII, HinP1I and Aci1I increased the coverage of CpGs sites in the probes designed for EDMA from 2.3% to 8.6% and genomic coverage from 4.8% to 6.1% (see Additional file 2: Table S2). This coverage accounts for a little more than 2.3 M CpG sites, which corresponds to about 10% of all CpG sites in the bovine genome. Similar coverage is obtained using reduced-representation bisulfite sequencing (RRBS), which has been shown to be efficient with small DNA input [19, 49, 50]. Both EDMA and RRBS have several steps in common, including genomic fragmentation by restriction enzyme, adaptor ligation and PCR amplification. The main benefits of reduced-representation bisulfite sequencing are that it allows quantitative evaluation of methylation at single-base resolution and is applicable to all species since it is based on DNA sequencing, which does not require a priori knowledge of the genome . However, EDMA similar to any other MSREs enrichment-based approaches generate a list of DMRs ranked in order of significance but does not provide information regarding the extent of DNA methylation which must be determined by targeted pyrosequencing. Furthermore, enrichment-based approaches using antibody or methyl-binding proteins have not been thoroughly tested with samples containing only a few ng of DNA. With larger DNA input, these methods have been shown to provide moderate resolution, since the capture depends on fragment methylation density [11, 52].
While epigenome-wide association studies and the development of technological platforms suitable for low input DNA are broadening in scope, improvement to the standardization of experimental assays across samples and to data analysis and interpretation remains slow . Although amplicons obtained by LM-PCR could be processed into sequencing libraries, we opted for microarray-based identification, which is more restrictive (being limited to the probes printed on the array) but has the major benefit of compatibility with an established, robust and user-friendly data analysis pipeline. These features may limit genomic coverage but definitely increase sample turnover rate by decreasing the time required to interpret data.
In comparison to gene expression, which is most often limited to the study of protein encoded transcribed elements which account for little more than 1% of the eukaryote genome, profiling overall DNA methylation considers a vastly more complex diversity of sequences, which in turn complicate data analysis. We therefore sought to support data interpretation by binning the information according to genomic features. Such an approach increases the statistical power to identify subtle alterations in genome regions by avoiding P value dilution, through multiple-testing corrections that include the vast majority of regions considered a priori unlikely to be differentially methylated . The data analysis pipeline built in to our platform accounts for: i) site specificity (e.g. promoter, intron, and exon), ii) region (e.g. shore, shelf, open-sea) and iii) sequence composition (e.g. CpG island density and length). This is in accordance with recent recommendations for epigenome-wide association studies .
The enrichment categorization implemented in the data analysis pipeline is based on previous studies [49, 55–60] and provides a full complement of graphic outputs. Even though the graphs might not all be relevant to all users, depending on the experimental design and biological hypothesis, are all produced automatically at no additional cost in time or resources on the part of the user. Overall, EDMA was developed as a cost-effective standardized platform that robustly profiles DNA methylation across the entire bovine genome. This sample processing platform is also complemented with bioinformatics supports for data analysis with the specific aim to aid data interpretation.
The importance of genome-scale parallel analysis of the DNA methylome and transcriptome
Current technological advances and the exponential growth of epigenetics studies in the past few years, in particular genome-wide studies, are advancing our knowledge and providing more evidence for the interdependence of epigenetic and genetic variations . By providing genome-wide parallel survey of the DNA methylome and the transcriptome for the same sample, EDMA offers a powerful tool for revealing highly relevant targets and potential associations between the DNA methylome and the transcriptome. In addition, overlaying methylomic and transcriptomic data has been shown to provide complementary information .
Genome-wide profiles of bovine sperm and blastocyst DNA methylomes
As proof of concept, we compared the DNA methylation profiles of bovine sperm and blastocysts and performed validation by pyrosequencing on selected DMRs. All selected candidates were found to be substantially more highly methylated in sperm DNA, corroborating the results obtained with EDMA. Physiologically relevant data were also generated and compared to the current literature. The sperm genome was found hyper-methylated compared to the embryo, which is consistent with the de-methylation process known to occur after fertilization in bovine  and other species such as mouse [49, 50] and zebra fish . The data also indicated that large numbers of loci methylated in bovine embryos are also methylated in sperm DNA, which also corroborates a previous report . As previously observed in mice , we found that from the DMRs enrichment, that short length/low density CpG islands showed a higher tendency to changes (i.e. being a DMR) than other regions. The direction of change is predominantly toward hypermethylation in sperm, but not always. In addition, we observed a high level of methylation in many repetitive elements, which is also in agreement with the patterns observed in mice [49, 50].
We observed marked differences (more hypermethylated DMRs in sperm) in the extent of methylation in long-terminal-repeat (LTR) retrotransposon, similar to previous findings with mice . Furthermore, it has been shown that the sperm contributed DMRs in pre-implantation embryos at LTRs were associated with reduction of DNA methylation and the most drastic methylation changes (reduction) in the sperm-to-zygote transition observed in some families of LINE and LTR retrotransposon . In our previous report, we also observed a transition in DNA methylation in LTR during bovine embryo development from Day 7 to Day 12 .
The reason for the observed large difference in LTR methylation during early development and, as shown specifically in this study, between sperm and blastocysts is not yet clear, although these differences might reflect the importance of de novo establishment of genome-wide methylation in the bovine early embryo. In this regard, robust LTR retrotransposon up-regulation has been shown during activation of the bovine embryo genome  and found to be a general requirement for progression through to the cleavage stage in mouse embryos [64, 65]. However, the nature of its role in mammalian early development remains elusive. Furthermore, it has been suggested that LTR re-methylation occurs in the early, rather than late, pre-implantation mouse embryo . Furthermore, the observation that a higher number of DMRs are present in low-complexity simple repeats in bovine blastocysts, in comparison with sperm, might represent notable dynamic changes in 5 mC in these specific class of repetitive elements during bovine early embryo development and differentiation .
By developing EDMA, we are providing a unified and reliable approach to analyze small amounts of genomic DNA, one that offers a good balance between genomic coverage and data turnaround time. The use of a microarray for fragment identification is robust, efficiently minimizing sample-to-data time. The built-in data analysis pipeline provides efficient means for data interpretation. The integrated data analysis pipeline could be a good option for researchers with limited bioinformatics resources. The platform is at the present time specific for the bovine genome, but a similar approach could be adapted easily to any species of which the entire genome is known. Such platforms enable the study of the potential epigenetic risks associated with assisted reproductive technologies (ART) or to highlight the sequence of events occurring during the establishment of embryonic cell lines.
Experiments took place in compliance with the guidelines of the Canadian Council on Animal Care and supervised by the Animal Protection Committee of Université Laval. These guidelines are strictly followed by the local abattoir and L’Alliance Boviteq who provided all the tissues and samples. The study did not require handling animals on university premises.
The design of the EmbryoGENE (http://embryogene.ca) DNA Methylation Array (EDMA) was based on a compilation of methylation-sensitive genetic loci found previously to be involved in early embryonic development . To maximize genomic coverage, all probes on the EDMA array were designed on the assumption that genomic cleaving using the MseI (T/TAA) restriction endonuclease would be nearly complete. All probes therefore targeted a specific MseI-MseI fragment within the bovine genome. Target loci were selected on the basis of their proximity to either the putative sites identified in our previous study  or to known CpG islands. Tiling of fragments neighboring the selected loci was then carried out until enough loci were selected to fill one Agilent SurePrint 1×1M slide (Agilent Technologies). Test hybridizations were carried out, and the best-performing 400,000 probes were selected for placement on the final Agilent SurePrint 2×400K array. Probe quality was measured by assessing sequence specificity and signal strength variation across the set of test hybridizations. The final EDMA array (EDMA.V2) contains 414,566 probes targeting 359,738 loci, surveying 20,355 genes and 34,379 CpG islands. The microarray also contains 10,388 control probes accounting for 2.5% of all the total probes, representing 5,610 Agilent proprietary spiked-in controls, 4,634 genomic cleavage controls and 144 EDMA spiked-in methylation controls. Controls were designed with an MseI restriction site at their center and were tiled at every 1 M base pairs throughout the bovine genome. These were used to assess the degree of genomic digestion. The EDMA spiked-in controls are exogenous DNA fragments (Solanum lycopersicum) chosen for their lack of homology with the bovine genome and for the presence of specific HpaII, AciI and HinP1I restriction sites within their sequence. They were artificially methylated or left unmethylated to provide positive and negative controls for the methylation-sensitive cleavage. Probe design for the EDMA microarray was carried out by Genotypic Inc. (Bangalore, India). The arrays were printed using the SurePrint technology (Agilent Technologies). The details of the EDMA platform array have been deposited in NCBI’s Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO, Platform accession number: GPL18384.
Sample production and genomic DNA extraction
Bovine blastocyst production in synthetic oviduct fluid (SOF) media was performed as described previously . The ovaries were collected at a local abattoir which is complying with the guidelines provided by the Canadian Council on Animal Care. Only expanded-blastocyst-stage embryos were collected, in four pools of ten embryos (n = 40). All embryos were washed three times with RNAse-free phosphate-buffered saline (PBS) prior to snap-freezing and storage at −80°C. Extraction of genomic DNA was carried out using an AllPrep DNA/RNA Mini Kit (QIAGEN, Mississauga, ON, Canada), and samples were eluted in 30 μL. Straws containing frozen semen from Holstein bulls were obtained from L’Alliance Boviteq Inc (Saint-Hyacinthe, QC, Canada). L’Alliance Boviteq Inc is a commercial service provider also complying with the guidelines provided by the Canadian Council on Animal Care. Sperm genomic DNA was extracted using lysis buffer, followed by ethanol precipitation. The quality and quantity of extracted DNA were analyzed by optical absorbance using a Nanodrop ND-1000 spectrophotometer (NanoDrop Technologies Inc.) and by electrophoresis on 0.5% agarose gel at 45 volts for 2 h.
The same DNA input (10 ng) was used for both embryo and sperm samples. For quality control purposes, all samples were spiked in with a mixture of six synthetic DNA constructs harbouring MseI restriction sites at each end. One pair of constructs was designed for each methyl-sensitive restriction enzyme (HpaII, HinP1I and Aci1I) to be used in the fragment selection process. For each pair, one control DNA fragment was methylated in vitro using CpG methyltransferase M.SssI (New England Biolabs). The difference between the extents of cleavage obtained with each pair was used to calculate cleavage efficiency. Residual MseI sites within the fragments were measured after processing to determine the extent of genomic fragmentation.
Sample (in 30 μL of elution buffer) plus 0.5 μL of bovine serum albumin (New England Biolabs), 5 μL of 10× Buffer 4 (New England Biolabs) and 28 μL of DNAse/RNAse free water were divided into two equal fractions and fragmented using 10 U of MseI (New England Biolabs) as follows, 16 h at 37°C followed by 65°C for 20 min. The two fractions were combined and DNA was concentrated by ethanol precipitation following addition of 5 μg of linear acrylamide solution (Ambion) as carrier and 5 μL of sodium acetate buffer (3 M, pH 5.5, Ambion Inc.). The pellet was washed twice with 70% ethanol, air-dried and resuspended in 5 μL of DNAse/RNAse free water.
Fragmented genome DNA adapter ligation
Adaptors were ligated to the MseI digested genomic fragments as described previously . Briefly, 5 μM of each primer (MseLig 21: 5′-AGT GGG ATT CCG CAT GCT AGT-3′, MseLig 12: 5′-TAA CTA GCA TGC-3′, IDT DNA), 0.5× One-Phor-All plus Buffer (Pharmacia Biotech) and 1.5 μL of nuclease-free water were added to the sample. Annealing was initiated at 65°C for 1 min, and the temperature was ramped at 1°C/min down to 15°C. T4 DNA ligase (5 units, Boehringer Mannheim) and 10 nmol of ATP were added and the reaction and incubated for 16 h at 15°C.
HpaIItiny fragment enrichment by ligation-mediated PCR (HELP) cocktail cleavage of ligated genomic DNA
The ligated sample was triple-cleaved with FastDigest™ methyl-sensitive restriction endonucleases (MSRE) HpaII (C/CGG), HinP1I (GC/GC), Aci1I (C/CGC) (Fermentas Thermo Fisher Scientific) using a sequential digestion. Samples were first digested in 50 μL reactions that included 10 μL of sample, 0.5 μL of HpaII, 0.5 μL of HinP1I, 5 μL of 10× FastDigest buffer and 34 μL of nuclease-free water for 12 h at 37°C. A second cleavage was performed by adding 0.5 μL of Aci1I, 5 μL of 10× FastDigest buffer and 44.5 μL of nuclease-free water and incubating at 37°C for 4 h, followed by thermal inactivation at 85°C for 10 min.
Verification of cleavage: qPCR of spiked-in templates
The extent of cleavage by MSREs was determined using qPCR detection of spiked-in controls. A master mixture (19 μL) containing 2 mM MgCl2, 1× LightCycler FastStart DNA Master SYBR Green mix (Roche Diagnostics Canada, Laval, QC, Canada), 14.4 μL of nuclease-free water and 1 μL of cleaved sample (template) was divided into three fractions to which 0.25 μM of each forward and reverse primer designed to target the appropriate control template was added (see Additional file 2: Table S1). Each qPCR run also included a positive control (non-digested spiked-in, 1/1000 dilution) and negative control (no template). The qPCR conditions were as follows: initial denaturation was carried out at 95°C for 10 minutes, followed by 50 amplification cycles at 95°C for 5 seconds, 52°C for 5 seconds and 72°C for 20 seconds. Melting curve analysis was performed for 1 cycle with a ramp rate of 0.2°C per second, starting at 94°C for 5 seconds, 72°C for 30 seconds, and back to 94°C for 0 seconds, and cooling at 40°C.
Amplicon specificity was determined from the shape of the melting curve and the difference in cycle threshold (Δ Ct) between methylated (MSRE cleavage protected) and unmethylated (MSRE cleavage unprotected) templates was used to calculate cleavage efficiency. The DNA samples were then precipitated by ethanol, washed and dissolved in 10 μL of nuclease-free water.
Fragment selection by ligation-mediated PCR
Selective amplification of methylated fragments was performed using two rounds of ligation-mediated PCR (LM-PCR) as described previously , with some modifications. The first LM-PCR amplification was carried out in 50 μL using 10 μL of sample to which 1 μM of primers (MseLig12 and MseLig21) was added, 0.1× One-Phor-All plus Buffer (Pharmacia Biotech), 0.6× Ex Taq™ buffer without Mg2+ (TaKaRa), 0.1 U Ex Taq™ enzyme (TaKaRa), 1.5 mM MgCl2 , 0.4 mM dNTP , and 23.5 μL of nuclease-free water. The mixture was then subjected to thermal cycling as follows: 94°C (40 sec), 57°C (30 sec) and 72°C (1 min 15 sec) for 15 cycles; 94°C (40 sec), 57°C (30 sec) and 72°C (1 min, 45 sec) for 34 cycles; and 94°C (40 sec), 57°C (30 sec) and 72°C (5 min) for the final cycle. The PCR products were resolved on 1% agarose gel to assess the quality.
To obtain sufficient DNA for downstream processing, three PCR reactions were carried out using 0.75 μL aliquots from the first LM PCR, to which 1.5 μM MseLig21 primer was added, plus 1× Buffer 1 (Roche/Boehringer Mannheim, Expand Long Template), 0.2 mM dNTPs, 0.8 μL of Expand Long Template Enzyme Mix (Roche/Boehringer Mannheim) and 38.7 μL of nuclease-free water. The following four-step programs were used: 94°C (60 sec), 65°C (30 sec), and 72°C (2 min) for 1 cycle; 94°C (40 sec), 65°C (30 sec) and 72°C (90 sec) for14 cycles; 94°C (40 sec), 65°C (30 sec) and 72°C (2 min) for 9 cycles and 72°C (5 min) as the final cycle. The quality and concentration of DNA were evaluated as described above.
The amplified product from the second LM-PCR was purified using a QIAquick PCR Purification Kit (Qiagen) according to the manufacturer’s instructions with minor modifications. Final elution volume was 41.5 μL, of which 1.5 μL was used for quantification by absorbance measurement. The rest was divided in two separate 20 μL aliquots. Adaptors were removed by digestion for 16 h at 37°C with 1 U of MseI (New England Biolabs). The mixture also contained 1× bovine serum albumin (New England Biolabs, Ipswich, MA, USA), 1 × Buffer 4 (New England Biolabs), and 24 μL of nuclease-free water. The reaction was terminating by heating for 20 min at 65°C. The sample was then purified using a QIAquick PCR Purification Kit (Qiagen), eluted in 23.5 μL of nuclease-free water and quantified by absorbance.
Sample labeling and hybridization
For each sample, 2.5 μg of DNA was labeled using the Universal Linkage System (ULS) labeling kit (Kreatech Biotechnology) according to the manufacturer’s instructions with minor modifications: 1 μL of Cy-ULS dye was added per 1 μg of genomic DNA adjusted with 10× labeling buffer. The labelling mixture was then held for 30 min at 85°C in a thermocycler, followed by 3 min on ice. Non-reacted ULS-Cy3/5 was removed by purification using a QIAquick PCR Purification Kit and samples were eluted in 23.5 μL nuclease-free water. A 1.5 μL aliquot was used to determine DNA concentration and dye incorporation using the ND-1000 NanoDrop. Hybridizations were performed according to the microarray manufacturer’s instructions (Agilent Technologies). Briefly, 1 μg of labeled sample (in 40 μL) was mixed with 158 μL of hybridization master mix containing 25 μL of bovine Cot-1 DNA (1.0 mg/mL, Bovine Hybloc competitor DNA, Applied Genetics Laboratories), 2.6 μL of Agilent 100x Blocking Agent and 130 μL of Agilent 2× HI-RPM Hybridization Buffer (Agilent Technologies). Samples were held at 95°C for 3 min and at 37°C for 30 min followed by addition of 65 μL of Agilent-CGHBlock (final volume 260 μL). The samples were loaded onto the microarray and hybridization was carried out in a hybridization oven (Shel Lab) for 40 h at 65°C and 20 rpm. Washing was carried out according to the microarray manufacturer’s instructions and slides were scanned with the PowerScanner (Tecan) and analyzed with Array-Pro Analyzer 6.3 software (MediaCybernetics).
Data analysis pipeline
To complement our transcriptomic platform , a complete suite of data analysis tools was added and integrated into the EmbryoGENE LIMS and Microarray Analysis (ELMA) gateway (http://elma.embryogene.ca/). This pipeline processes sample-associated information, experimental design and protocols, as well as microarray data analysis for identification of differential gene/loci lists and further data mining for enrichment analyses. EDMA data was analyzed using the Limma package from Bioconductor [67, 68]. First, Loess intra-array normalization followed by quantile inter-array scale normalization were applied. Normalized data was then fitted to a linear model and Bayesian statistics of differential expression were obtained. Differences in DNA methylation were considered significant when the P value was < 0.05 and the absolute log2 fold-change was at least 1.5. The inter-treatment comparisons were conducted using between-group analysis based on Eigen values using the Bioconductor package MADE4 . Visualization tools were developed to generate plots of the extent of cleavage of spiked-in controls and overall genomic cleavage for quality control purposes. Different plots including Volcano plots were generated to visualize the amount of differentially methylated regions among treatments. Normalized data and a list of differentially methylated loci can be exported as text files for downstream data mining. Different enrichment analyses for genome-scale DNA methylation data are performed through a string of integrated scripts that categorize the information based on CpG island density , CpG island length , CpG island distance , genomic location and types of repetitive elements . The pipeline also pinpoints methylation hot spots, generates bedgraph files appropriate for visualization within a genome browser, and uses Circos plot  to generate a circular graph representing overall methylation levels and correlating these to transcription levels, if the latter data are available.
The CpG island lengths, density and positions for the UMD3.1 (ftp://ftp.cbcb.umd.edu/pub/data/assembly/Bos_taurus/) build of the bovine genome was obtained from the UCSC genome browser . For any given gene in EDMA, there are five types of annotation “windows” including; 1) Distal promoter, 2) Promoter, 3) Proximal Promoter, 4) Exons and 5) Introns. The “Proximal Promoter”, “Promoter” and “Distal Promoter” regions are defined as the first 1 kbp, 5 kbp and 50 kbp 5′ of the transcription start site (TSS). For each probes, genes are added to the appropriate columns if their MseI-MseI fragment overlap those windows. Genomic locations which were not part of a CpG island were further split into three types based on their distance from the nearest CpG island: “CpG shores” for regions within 2,000 nucleotides of an island, “CpG shelves” for regions between 2,000 and 4,000 nucleotides of an island and “Open Sea” for regions further away . Probes whose fragment overlaps a CpG island were annotated using the descriptive characteristics of that island. Of all CpG island surveyed by the array, those in the bottom 20 percentile of length were classified as “short”, those at the top 80 percentile were classified as “long” and all others were classified as “intermediate” . The same scheme was applied to CpG island density, with the bottom 20 percentile being labelled “low density”, the top 80 percentile being labelled “high density”, and all others being labelled “Intermediate density” . Bovine repetitive elements content were identified by RepeatMasker (http://www.repeatmasker.org/) with build 20120418 of the RepBase database  and then used the repeat classes attribute (LINE, SINE, etc.) as a basis for category enrichment. The “methylation hot spots” is calculated as the averages p-values of differential methylation over windows of 100 K nucleotides. More specifically, for all probes on the array, we look up all other probes within 100 K nucleotides upstream and downstream, and average the p-values thus obtained. The averaged p-values have no statistical meaning, but can be used as an indicator for regions of interest, which we called a “methylation hot spot”.
Pyrosequencing was performed as described previously . Briefly, blastocyst and sperm DNA samples were bisulfite-converted using an EZ DNA methylation direct KIT, (ZYMO Research) followed by amplification of target loci by PCR prior to sequencing. For all pyrosequencing assays, PyroMark Assay Design software (Qiagen) was used to design three oligonucleotide primers (forward, reverse and sequencing), synthesized by Integrated DNA technologies (Coralville). All reverse primers were biotinylated at the 5′ terminus and purified by HPLC. PCR amplification was carried out in 25 μL containing 0.2 μM of each forward and reverse 5′-biotinylated primer, 1.25 U of Platinum Taq DNA polymerase (Invitrogen), 1× Taq DNA polymerase buffer, 0.2 mM dNTPs, 3–4 mM MgCl2, 2 μL of bisulfite-treated DNA and 17.5 μL of nuclease-free water. The following three-step program was used: 95°C (5 min) for 1 cycle, 95°C (30 sec), 48°C for DMR1b, 49°C for DMRs 2 and 3a, 50°C for DMRs 1a, 3b, 5–8 and 54°C for DMR 4 (30 sec) and 72°C (30 sec) for 35 cycles, and finally 72°C (5 min). The specificity of the amplification was verified by electrophoresis on 2% w/v agarose gel at 90 V for 45 min. Templates were purified by adding 2 μL of streptavidin-coated Sepharose beads (GE Healthcare) in the presence of 40 μL of binding buffer (Qiagen) and then pyrosequenced . Pyrosequencing reactions were conducted using a PyroMark Q24 apparatus (Qiagen).
Availability of supporting data
The data analysis pipeline user manual is available at; [EDMA user manual]: http://emb-bioinfo.fsaa.ulaval.ca/bioinfo/html/epigenetics/Epigenetics%20Analysis%20Pipeline.pdf.
The dataset of microarray results have been deposited in NCBI’s Gene Expression Omnibus (GEO) and are accessible through GEO series accession number: GSE57709 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE57709).
Assisted reproductive technologies
Differentially methylated region
EmbryoGENE DNA methylation analysis
EmbryoGENE LIMS and microarray analysis
HpaII tiny fragment enrichment by ligation-mediated PCR
Long interspersed element
Long terminal repeat retrotransposons
Methyl-sensitive restriction endonuclease
Paternally expressed 10 gene
Reduced-representation bisulfite sequencing
Short interspersed element.
Shojaei Saadi H. A. supported by EmbryoGENE, REDIH and RQR-CREATE scholarships. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (grant number: NETGP-340825-06). This grant was awarded from the Strategic Research Network program supporting the EmbryoGENE Network (http://www.embryogene.ca).
- Robert C, Nieminen J, Dufort I, Gagné D, Grant J, Cagnone G, Plourde D, Nivet A-L, Fournier É, Paquet É, Blazejczyk M, Rigault P, Juge N, Sirard MA: Combining resources to obtain a comprehensive survey of the bovine embryo transcriptome through deep sequencing and microarrays. Mol Reprod Dev. 2011, 78 (9): 651-664.PubMedView ArticleGoogle Scholar
- Tsoi S, Zhou C, Grant J, Pasternak J, Dobrinsky J, Rigault P, Nieminen J, Sirard M-A, Robert C, Foxcroft G, Dyck MK: Development of a porcine (Sus scofa) embryo-specific microarray: array annotation and validation. BMC Genomics. 2012, 13 (1): 370-PubMed CentralPubMedView ArticleGoogle Scholar
- Gad A, Hoelker M, Besenfelder U, Havlicek V, Cinar U, Rings F, Held E, Dufort I, Sirard MA, Schellander K, Tesfaye D: Molecular mechanisms and pathways involved in bovine embryonic genome activation and their regulation by alternative in vivo and in vitro culture conditions. Biol Reprod. 2012, 87 (4): 100-PubMedView ArticleGoogle Scholar
- Cagnone GLM, Sirard M-A: Transcriptomic signature to oxidative stress exposure at the time of embryonic genome activation in bovine blastocysts. Mol Reprod Dev. 2013, 80 (4): 297-314.PubMedView ArticleGoogle Scholar
- Cagnone GL, Dufort I, Vigneault C, Sirard MA: Differential gene expression profile in bovine blastocysts resulting from hyperglycemia exposure during early cleavage stages. Biol Reprod. 2012, 86 (2): 50-PubMedView ArticleGoogle Scholar
- Zhang P, Zucchelli M, Bruce S, Hambiliki F, Stavreus-Evers A, Levkov L, Skottman H, Kerkelä E, Kere J, Hovatta O: Transcriptome profiling of human pre-implantation development. PLoS One. 2009, 4 (11): e7844-PubMed CentralPubMedView ArticleGoogle Scholar
- Hamatani T, Carter M, Sharov A, Ko M: Dynamics of global gene expression changes during mouse preimplantation development. Dev Cell. 2004, 6 (1): 117-131.PubMedView ArticleGoogle Scholar
- Léandri R, Archilla C, Bui L, Peynot N, Liu Z, Cabau C, Chastellier A, Renard J, Duranthon V: Revealing the dynamics of gene expression during embryonic genome activation and first differentiation in the rabbit embryo with a dedicated array screening. Physiol Genomics. 2009, 36 (2): 98-113.PubMedView ArticleGoogle Scholar
- Szyf M: The early life environment and the epigenome. Biochim Biophys Acta. 2009, 1790 (9): 878-885.PubMedView ArticleGoogle Scholar
- Guerrero-Bosagna C, Skinner M: Environmentally induced epigenetic transgenerational inheritance of phenotype and disease. Mol Cell Endocrinol. 2011, 354 ((1-2)): 3-8.PubMed CentralPubMedGoogle Scholar
- Suzuki MM, Bird A: DNA methylation landscapes: provocative insights from epigenomics. Nat Rev Genet. 2008, 9 (6): 465-476.PubMedView ArticleGoogle Scholar
- Bernstein B, Meissner A, Lander E: The mammalian epigenome. Cell. 2007, 128 (4): 669-681.PubMedView ArticleGoogle Scholar
- Baccarelli A, Rienstra M, Benjamin EJ: Cardiovascular epigenetics: basic concepts and results from animal and human studies. Circ Cardiovasc Genet. 2010, 3 (6): 567-573.PubMed CentralPubMedView ArticleGoogle Scholar
- Shi L, Wu J: Epigenetic regulation in mammalian preimplantation embryo development. Reprod Biol Endocrinol. 2009, 7 (1): 59-PubMed CentralPubMedView ArticleGoogle Scholar
- Bird A: DNA methylation patterns and epigenetic memory. Genes Dev. 2002, 16 (1): 6-21.PubMedView ArticleGoogle Scholar
- Raynal N, Si J, Taby R, Gharibyan V, Ahmed S, Jelinek J, Estécio M, Issa J-PJ: DNA methylation does not stably lock gene expression but instead serves as a molecular mark for gene silencing memory. Cancer Res. 2012, 72 (5): 1170-1181.PubMed CentralPubMedView ArticleGoogle Scholar
- Talens RP, Boomsma DI, Tobi EW, Kremer D, Jukema JW, Willemsen G, Putter H, Slagboom PE, Heijmans BT: Variation, patterns, and temporal stability of DNA methylation: considerations for epigenetic epidemiology. FASEB J. 2010, 24 (9): 3135-3144.PubMedView ArticleGoogle Scholar
- Fraga MF, Ballestar E, Paz MF, Ropero S, Setien F, Ballestar ML, Heine-Suñer D, Cigudosa JC, Urioste M, Benitez J, Boix-Chornet M, Sanchez-Aguilera A, Ling C, Carlsson E, Poulsen P, Vaag A, Stephan Z, Spector TD, Wu YZ, Plass C, Esteller M: Epigenetic differences arise during the lifetime of monozygotic twins. Proc Natl Acad Sci USA. 2005, 102 (30): 10604-10609.PubMed CentralPubMedView ArticleGoogle Scholar
- Wu H, Zhang Y: Early embryos reprogram DNA methylation in two steps. Cell Stem Cell. 2012, 10 (5): 487-489.PubMedView ArticleGoogle Scholar
- Seisenberger S, Peat JR, Reik W: Conceptual links between DNA methylation reprogramming in the early embryo and primordial germ cells. Curr Opin Cell Biol. 2013, 25 (3): 281-288.PubMedView ArticleGoogle Scholar
- Reik W, Dean W, Walter J: Epigenetic reprogramming in mammalian development. Science. 2001, 293 (5532): 1089-1093.PubMedView ArticleGoogle Scholar
- Morgan HD, Santos F, Green K, Dean W, Reik W: Epigenetic reprogramming in mammals. Hum Mol Genet. 2005, 14 (suppl 1): R47-R58.PubMedView ArticleGoogle Scholar
- Seisenberger S, Peat JR, Hore TA, Santos F, Dean W, Reik W: Reprogramming DNA methylation in the mammalian life cycle: building and breaking epigenetic barriers. Philos Trans R Soc B Biol Sci. 2013, 368 (1609): 20110330-View ArticleGoogle Scholar
- Santos F, Hendrich B, Reik W, Dean W: Dynamic Reprogramming of DNA Methylation in the Early Mouse Embryo. Dev Biol. 2002, 241 (1): 172-182.PubMedView ArticleGoogle Scholar
- van Montfoort APA, Hanssen LLP, de Sutter P, Viville S, Geraedts JPM, de Boer P: Assisted reproduction treatment and epigenetic inheritance. Hum Reprod Update. 2012, 18 (2): 171-197.PubMed CentralPubMedView ArticleGoogle Scholar
- McGraw S, Shojaei Saadi HA, Robert C: Meeting the methodological challenges in molecular mapping of the embryonic epigenome. Mol Hum Reprod. 2013, 19 (12): 809-827.PubMedView ArticleGoogle Scholar
- de Montera B, Fournier E, Shojaei Saadi HA, Gagne D, Laflamme I, Blondin P, Sirard MA, Robert C: Combined methylation mapping of 5mC and 5hmC during early embryonic stages in bovine. BMC Genomics. 2013, 14: 406-PubMed CentralPubMedView ArticleGoogle Scholar
- Irizarry RA, Ladd-Acosta C, Carvalho B, Wu H, Brandenburg SA, Jeddeloh JA, Wen B, Feinberg AP: Comprehensive high-throughput arrays for relative methylation (CHARM). Genome Res. 2008, 18 (5): 780-790.PubMed CentralPubMedView ArticleGoogle Scholar
- Slomko H, Heo HJ, Einstein FH: Minireview: epigenetics of obesity and diabetes in humans. Endocrinology. 2012, 153 (3): 1025-1030.PubMed CentralPubMedView ArticleGoogle Scholar
- Youngson NA, Morris MJ: What obesity research tells us about epigenetic mechanisms. Philos Trans R Soc B Biol Sci. 2013, 368 (1609): 20110337-View ArticleGoogle Scholar
- Saab BJ, Mansuy IM: Neurobiological disease etiology and inheritance: an epigenetic perspective. J Exp Biol. 2014, 217 (1): 94-101.PubMedView ArticleGoogle Scholar
- Tsankova N, Renthal W, Kumar A, Nestler EJ: Epigenetic regulation in psychiatric disorders. Nat Rev Neurosci. 2007, 8 (5): 355-367.PubMedView ArticleGoogle Scholar
- Iacobuzio-Donahue CA: Epigenetic changes in cancer. Annu Rev Pathol Mech Dis. 2009, 4 (1): 229-249.View ArticleGoogle Scholar
- Dawson Mark A, Kouzarides T: Cancer epigenetics: from mechanism to therapy. Cell. 2012, 150 (1): 12-27.PubMedView ArticleGoogle Scholar
- Rodriguez-Paredes M, Esteller M: Cancer epigenetics reaches mainstream oncology. Nat Med. 2011, 17 (3): 330-339.PubMedView ArticleGoogle Scholar
- Feinberg AP: Phenotypic plasticity and the epigenetics of human disease. Nature. 2007, 447 (7143): 433-440.PubMedView ArticleGoogle Scholar
- Kantlehner M, Kirchner R, Hartmann P, Ellwart JW, Alunni-Fabbroni M, Schumacher A: A high-throughput DNA methylation analysis of a single cell. Nucleic Acids Res. 2011, 39 (7): e44-PubMed CentralPubMedView ArticleGoogle Scholar
- Denomme MM, Zhang L, Mann MR: Single oocyte bisulfite mutagenesis. J Vis Exp. 2012, e4046-64
- Lorthongpanich C, Cheow LF, Balu S, Quake SR, Knowles BB, Burkholder WF, Solter D, Messerschmidt DM: Single-cell DNA-methylation analysis reveals epigenetic chimerism in preimplantation embryos. Science. 2013, 341 (6150): 1110-1112.PubMedView ArticleGoogle Scholar
- Klein C, Schmidt-Kittler O, Schardt J, Pantel K, Speicher M, Riethmüller G: Comparative genomic hybridization, loss of heterozygosity, and DNA sequence analysis of single cells. Proc Natl Acad Sci USA. 1999, 96 (8): 4494-4499.PubMed CentralPubMedView ArticleGoogle Scholar
- Oda M, Glass JL, Thompson RF, Mo Y, Olivier EN, Figueroa ME, Selzer RR, Richmond TA, Zhang X, Dannenberg L, Green RD, Melnick A, Hatchwell E, Bouhassira EE, Verma A, Suzuki M, Greally JM: High-resolution genome-wide cytosine methylation profiling with simultaneous copy number analysis and optimization for limited cell numbers. Nucleic Acids Res. 2009, 37 (12): 3829-3839.PubMed CentralPubMedView ArticleGoogle Scholar
- Hashimoto K, Kokubun S, Itoi E, Roach HI: Improved quantification of DNA methylation using methylation-sensitive restriction enzymes and real-time PCR. Epigenetics. 2007, 2 (2): 86-91.PubMedView ArticleGoogle Scholar
- Laird P: Principles and challenges of genomewide DNA methylation analysis. Nat Rev Genet. 2010, 11 (3): 191-203.PubMedView ArticleGoogle Scholar
- Schumacher A, Kapranov P, Kaminsky Z, Flanagan J, Assadzadeh A, Yau P, Virtanen C, Winegarden N, Cheng J, Gingeras T, Petronis A: Microarray-based DNA methylation profiling: technology and applications. Nucleic Acids. 2006, 34 (2): 528-542.View ArticleGoogle Scholar
- Cross SH, Charlton JA, Nan X, Bird AP: Purification of CpG islands using a methylated DNA binding column. Nat Genet. 1994, 6 (3): 236-244.PubMedView ArticleGoogle Scholar
- Rauch T, Li H, Wu X, Pfeifer GP: MIRA-assisted microarray analysis, a New technology for the determination of DNA methylation patterns, identifies frequent methylation of homeodomain-containing genes in lung cancer cells. Cancer Res. 2006, 66 (16): 7939-7947.PubMedView ArticleGoogle Scholar
- Jones PA: Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat Rev Genet. 2012, 13 (7): 484-492.PubMedView ArticleGoogle Scholar
- Zilberman D, Henikoff S: Genome-wide analysis of DNA methylation patterns. Development. 2007, 134 (22): 3959-3965.PubMedView ArticleGoogle Scholar
- Smith Z, Chan M, Mikkelsen T, Gu H, Gnirke A, Regev A, Meissner A: A unique regulatory phase of DNA methylation in the early mammalian embryo. Nature. 2012, 484 (7394): 339-344.PubMed CentralPubMedView ArticleGoogle Scholar
- Smallwood SA, Tomizawa S, Krueger F, Ruf N, Carli N, Segonds-Pichon A, Sato S, Hata K, Andrews SR, Kelsey G: Dynamic CpG island methylation landscape in oocytes and preimplantation embryos. Nat Genet. 2011, 43 (8): 811-814.PubMed CentralPubMedView ArticleGoogle Scholar
- Harris RA, Wang T, Coarfa C, Nagarajan RP, Hong C, Downey SL, Johnson BE, Fouse SD, Delaney A, Zhao Y, Olshen A, Ballinger T, Zhou X, Forsberg KJ, Gu J, Echipare L, O'Geen H, Lister R, Pelizzola M, Xi Y, Epstein CB, Bernstein BE, Hawkins RD, Ren B, Chung WY, Gu H, Bock C, Gnirke A, Zhang MQ, Haussler D, et al: Comparison of sequencing-based methods to profile DNA methylation and identification of monoallelic epigenetic modifications. Nat Biotechnol. 2010, 28 (10): 1097-1105.PubMed CentralPubMedView ArticleGoogle Scholar
- Fouse SD, Nagarajan RO, Costello JF: Genome-scale DNA methylation analysis. Epigenomics. 2010, 2 (1): 105-117.PubMed CentralPubMedView ArticleGoogle Scholar
- Michels KB, Binder AM, Dedeurwaerder S, Epstein CB, Greally JM, Gut I, Houseman EA, Izzi B, Kelsey KT, Meissner A, Milosavljevic A, Siegmund KD, Bock C, Irizarry RA: Recommendations for the design and analysis of epigenome-wide association studies. Nat Methods. 2013, 10 (10): 949-955.PubMedView ArticleGoogle Scholar
- Gu H, Bock C, Mikkelsen TS, Jager N, Smith ZD, Tomazou E, Gnirke A, Lander ES, Meissner A: Genome-scale DNA methylation mapping of clinical samples at single-nucleotide resolution. Nat Methods. 2010, 7 (2): 133-136.PubMed CentralPubMedView ArticleGoogle Scholar
- Meissner A, Mikkelsen TS, Gu H, Wernig M, Hanna J, Sivachenko A, Zhang X, Bernstein BE, Nusbaum C, Jaffe DB, Gnirke A, Jaenisch R, Lander ES: Genome-scale DNA methylation maps of pluripotent and differentiated cells. Nature. 2008, 454 (7205): 766-770.PubMed CentralPubMedGoogle Scholar
- Weber M, Hellmann I, Stadler MB, Ramos L, Paabo S, Rebhan M, Schubeler D: Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome. Nat Genet. 2007, 39 (4): 457-466.PubMedView ArticleGoogle Scholar
- Elango N, Yi SV: Functional relevance of CpG island length for regulation of gene expression. Genetics. 2011, 187 (4): 1077-1083.PubMed CentralPubMedView ArticleGoogle Scholar
- Krausz C, Sandoval J, Sayols S, Chianese C, Giachini C, Heyn H, Esteller M: Novel insights into DNA methylation features in spermatozoa: stability and peculiarities. PLoS One. 2012, 7 (10): e44479-PubMed CentralPubMedView ArticleGoogle Scholar
- Doi A, Park IH, Wen B, Murakami P, Aryee MJ, Irizarry R, Herb B, Ladd-Acosta C, Rho J, Loewer S, Miller J, Schlaeger T, Daley GQ, Feinberg AP: Differential methylation of tissue- and cancer-specific CpG island shores distinguishes human induced pluripotent stem cells, embryonic stem cells and fibroblasts. Nat Genet. 2009, 41 (12): 1350-1353.PubMed CentralPubMedView ArticleGoogle Scholar
- Bock C, Beerman I, Lien WH, Smith ZD, Gu H, Boyle P, Gnirke A, Fuchs E, Rossi DJ, Meissner A: DNA methylation dynamics during in vivo differentiation of blood and skin stem cells. Mol Cell. 2012, 47 (4): 633-647.PubMed CentralPubMedView ArticleGoogle Scholar
- Dean W, Santos F, Stojkovic M, Zakhartchenko V, Walter J, Wolf E, Reik W: Conservation of methylation reprogramming in mammalian development: Aberrant reprogramming in cloned embryos. Proc Natl Acad Sci. 2001, 98 (24): 13734-13738.PubMed CentralPubMedView ArticleGoogle Scholar
- Andersen I, Reiner A, Aanes H, Alestrom P, Collas P: Developmental features of DNA methylation during activation of the embryonic zebrafish genome. Genome Biol. 2012, 13 (7): R65-PubMed CentralPubMedView ArticleGoogle Scholar
- Bui LC, Evsikov AV, Khan DR, Archilla C, Peynot N, Hénaut A, Le Bourhis D, Vignon X, Renard JP, Duranthon V: Retrotransposon expression as a defining event of genome reprograming in fertilized and cloned bovine embryos. Reproduction. 2009, 138 (2): 289-299.PubMedView ArticleGoogle Scholar
- Peaston AE, Evsikov AV, Graber JH, de Vries WN, Holbrook AE, Solter D, Knowles BB: Retrotransposons regulate host genes in mouse oocytes and preimplantation embryos. Dev Cell. 2004, 7 (4): 597-606.PubMedView ArticleGoogle Scholar
- Kigami D, Minami N, Takayama H, Imai H: MuERV-L is one of the earliest transcribed genes in mouse one-cell embryos. Biol Reprod. 2003, 68 (2): 651-654.PubMedView ArticleGoogle Scholar
- Kim M, Park Y-K, Kang T-W, Lee S-H, Rhee Y-H, Park J-L, Kim HJ, Lee D, Lee D, Kim S-Y, Kim YS: Dynamic changes in DNA methylation and hydroxymethylation when hES cells undergo differentiation toward a neuronal lineage. Hum Mol Genet. 2013, 23 (3): 657-667.PubMedView ArticleGoogle Scholar
- Smyth GK, Michaud J, Scott HS: Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics. 2005, 21 (9): 2067-2075.PubMedView ArticleGoogle Scholar
- McCarthy DJ, Smyth GK: Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics. 2009, 25 (6): 765-771.PubMed CentralPubMedView ArticleGoogle Scholar
- Culhane AC, Thioulouse J, Perriere G, Higgins DG: MADE4: an R package for multivariate analysis of gene expression data. Bioinformatics. 2005, 21 (11): 2789-2790.PubMedView ArticleGoogle Scholar
- Jurka J, Kapitonov VV, Pavlicek A, Klonowski P, Kohany O, Walichiewicz J: Repbase Update, a database of eukaryotic repetitive elements. Cytogenet Genome Res. 2005, 110 (1–4): 462-467.PubMedView ArticleGoogle Scholar
- Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, Jones SJ, Marra MA: Circos: an information aesthetic for comparative genomics. Genome Res. 2009, 19 (9): 1639-1645.PubMed CentralPubMedView ArticleGoogle Scholar
- Meyer LR, Zweig AS, Hinrichs AS, Karolchik D, Kuhn RM, Wong M, Sloan CA, Rosenbloom KR, Roe G, Rhead B, Raney BJ, Pohl A, Malladi VS, Li CH, Lee BT, Learned K, Kirkup V, Hsu F, Heitner S, Harte RA, Haeussler M, Guruvadoo L, Goldman M, Giardine BM, Fujita PA, Dreszer TR, Diekhans M, Cline MS, Clawson H, Barber GP, et al: The UCSC Genome Browser database: extensions and updates 2013. Nucleic Acids Res. 2013, 41 (D1): D64-D69.PubMed CentralPubMedView ArticleGoogle Scholar
- O’Doherty AM, O’Shea LC, Fair T: Bovine DNA methylation imprints are established in an oocyte size-specific manner, which are coordinated with the expression of the DNMT3 family proteins. Biol Reprod. 2012, 86 (3): 67-PubMedView ArticleGoogle Scholar
- Tost J, Gut IG: DNA methylation analysis by pyrosequencing. Nat Protoc. 2007, 2 (9): 2265-2275.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.